At the present time, massive amounts of data are being collected. The availability of such data gives rise to the urgent need to transform the data into knowledge; this is the function of the field of Knowledge Discovery in Database (KDD). The most essential step in KDD is the Data Mining (DM) step which is the search engine to find the knowledge embedded in the data. The tasks of DM can be classified into two types, namely: predictive or descriptive, according to the sought functionality. One of the older and well-studied functionalities in data mining is cluster analysis (Clustering). Clustering methods can be either hierarchal or partitioning. One of the very well known clustering algorithms is the C-means. In this paper, we turn our focus on cluster analysis in general and on the C-means partitioning method in particular. We direct our attention to the modification of the C-means algorithm in the way it calculates the means of the clusters. We consider the mean of a cluster to be one of the objects instead of being an imaginary point in the cluster. Our modified C-means (MC-means) algorithm is implemented in a system developed in the visual basic.net programming language. The well-known Soybean dataset is used in an experiment to evaluate our modification to the C-means algorithm. This paper is concluded with an analysis and discussion of the experiments’ result on the bases of several criteria.